> > > So actually an equality test against INT_MAX would be correct. But making
> > > that clear in the comment would probably not be a bad idea :-)
> >
> > I have applied a fix for this, because it obviously needed fixing
> > regardless of if it fixes the original issue all the way. Still looking
> > for confirmation if it does, though.
>
> The fix works well. I tested it with UTF-8 encoding and Japanese_Japan.932
> (SJIS) locale and I didn't see any errors. I think it it is enough to cover
> the buggy UTF-8 treatment on Windows.
>
Thanks for testing.
Bruce, that means the original patch is not needed, so you can take it off the patch queue. Thanks.
/Magnus